For serious infections due to vancomycin resistant Enterococcus faecium that are also highly ampicillin resistant, there is no known, effective therapy. Surprisingly little is known about E. faecium; indeed, apart from genes from mobile elements or for antibiotic resistances, there are only 5 E. faecium gene sequences deposited in GenBank. Also, there is no grant listed in the NIH databases on E. faecium. The general goals of this proposal are to increase our understanding of E. faecium and to identify properties possibly important for disease. Aim I is to identify in vivo expressed E. faecium proteins, to be accomplished by screening E. faecium DNA libraries with sera from humans with E. faecium infections (particularly endocarditis). The investigators have successfully utilized this technique in their E. faecalis research. The E. faecium libraries will also be screened with DNA probes from high interest E. faecalis genes, using low stringency hybridization, to look for possible homologs. (An alternative approach for enhancing for genes possibly involved in virulence would be to identify genes encoding export domains by generating alkaline phosphatase fusions.) Aim II is to identify and sequence genes encoding the antigens. This will be done by (1) subcloning DNase I generated fragments into expression vectors, selecting for immunopositive subclones, then sequencing the subcloned regions, or by (2) creating knock-out mutations of immunoreactive clones using a defective transposon and sequencing from the transposon ends; they have used both methods in their E. faecalis project. (If the alternative approach will PhoA fusions is used, inserts in blue colonies would be sequenced using a phoA primer.) Database searches will be performed to identify possible homologies. Aim III is to generate targeted knock-out mutations in E. faecium, starting with clones encoding homologs of proteins implicated in virulence of other organisms. Modified DNA will be introduced back into the E. faecium chromosome, by homologous recombination, and mutants tested in vitro and in vivo. The strategy is based on the hypotheses that (1) among in vivo expressed antigens and exported proteins are ones which are involved in production of disease, and that (2) preventative or therapeutic modalities can be derived from knowledge about these proteins. It is envisioned that results from this project will provide a foundation for the development of non-antibiotic methods (e.g., passive or active immunization) to prevent, control, or combat E. faecium infections, and may identify new targets for traditional antibiotic development strategies.